Markov chain Monte Carlo and acceptance–rejection algorithms for synthesising short‐term variations in the generation output of the photovoltaic system

Abstract

Solar photovoltaic (PV) is fast being deployed in electric power grids. For its more reliable integration into grids, the impact of its intermittent generation on grids should be further examined. However, high-resolution solar and meteorological measurement data in 1 s or 1 min intervals are often not available for many areas or they are expensive to obtain. Thus, the objective of this study is to present a first-order Markov chain Monte Carlo (MCMC) method that synthesises the short-term variation, or sudden energy shortages and overages, in the power generation of a PV system in high-resolution, from low-resolution solar and meteorological data. In addition, up and down ramp rates in PV output are significant for the application of battery storage systems or grid operation. Thus, the acceptance–rejection algorithm able to limit such ramp rates is proposed. The proposed MCMC and acceptance–rejection methods are verified by comparing the statistical characteristics of the synthesised data to those of the original data. From case studies, this study found that the proposed methods could present the sufficient short-term variation in generation output of a PV system.